Sitemap
A list of all the posts and pages found on the site. For you robots out there is an XML version available for digesting as well.
Pages
Posts
portfolio
Disease Prediction Model
An Analysis of Lassa Fever Outbreaks in Nigeria using Machine Learning Models and Shapley Values
Infectious Disease Dashboard
A Real-Time Data-Driven Infectious Disease Dashboard to Monitor 2022-2023 Global Mpox Outbreak
publications
Real-Time Hybrid Dashboard and App for Mpox Outbreak Surveillance
Published in IEEE Global Humanitarian Technology Conference (GHTC), 2023
With the continuous rise in the global threat of infectious diseases, surveillance systems have become powerful tools for monitoring the development and transmission patterns of fast-changing disease outbreaks, by public health officers. These surveillance systems typically track relevant epidemiological data such as number of cases, fatality rates, hot spots, and in most cases, evaluate the impact of public health intervention strategies in these locations. Data from these surveillance apps can also be used for efficient emergency response preparatory logistics, as well as early containment. In this paper, we presents a modular, open-source, GIS-enabled, web-based dashboard and mobile app that are capable of visualizing and identifying epidemiological patterns for any infectious disease of interest, in real time. These visualizations can be in the form of GIS maps, charts, and other metrics to track certain indicators. Python and Streamlit library are used to manage the frontend of the application. Data from the 2022 Mpox outbreak in the United States were used to evaluate the application. Nongovernmental organizations or other community-based groups can leverage on, and adapt this dashboard to monitor the spread of any quick-onset disease outbreak in their regions.
Recommended citation: D. Quezada, S. Akwafuo, and A. Wattamwar, "Real-Time Hybrid Dashboard and App for Mpox Outbreak Surveillance," presented at the 2023 IEEE Global Humanitarian Technology Conference (GHTC), Radnor, PA, USA, 2023, pp. 433-439, doi: [10.1109/GHTC56179.2023.10355026](https://doi.org/10.1109/GHTC56179.2023.10355026). http://danielquezada.com/files/mpox.pdf
Harnessing Machine Learning for Predictive Analytics: A Case Study of Lassa Fever Outbreaks in Nigeria
Published in IEEE International Conference on Control, Decision and Information Technologies, 2024
In the ongoing battle against global pandemics, understanding the key determinants that fuel outbreaks are of paramount importance. With this focus, our study aims to assess and rank the predictive capabilities of a wide range of socio-economic, eco-climatic, and spatiotemporal variables in predicting Lassa Fever (LF) outbreaks, using data from previous Nigerian outbreaks (2012-2019). Employing machine learning methods, particularly XGBoost and Random Forest, our study aims to offer accurate and robust predictions concerning LF incidence rates. As a crucial add-on, we leverage the innovative SHAP (SHapley Additive exPlanations) technique as a post-processing tool to dissect and better understand the contributions of individual features towards the predictions generated by our machine learning models. This multi-layered approach allowed us to place a pronounced focus on healthcare infrastructure, population demographics, land cover, and other climatic covariates. Among the models evaluated, XGBoost performed the best; delivering an accuracy of 0.93, and AUC of 0.90, and an F1 score of 0.86 on 2018 data. For 2019 data, it maintained a strong accuracy of 0.90, an AUC of 0.89, and an F1 score of 0.75. Our SHAP analysis further emphasized precipitation seasonality, diagnostic center density, and land cover characteristics as pivotal influencers in predicting LF outbreaks. These findings shed light on the complex interplay between environmental conditions, urbanization, and healthcare infrastructure. Given these promising results, our work sets the stage for the development of an advanced early warning system for Lassa Fever in Nigeria: a system that could efficiently intertwine computational insights with on-ground interventions, ensuring timely and targeted responses to potential outbreaks.
Recommended citation: D. Quezada, S. Akwafuo, and S. Halyal, "Harnessing Machine Learning for Predictive Analytics: A Case Study of Lassa Fever Outbreaks in Nigeria" to be presented at the 2024 IEEE International Conference on Control, Decision, and Information Technologies, Valletta, Malta, 2024 http://danielquezada.com/files/lassa.pdf